Can Collaborative Learning Be Private, Robust and Scalable?

Usynin D, Klause H, Paetzold JC, Rueckert D, Kaissis G (2022)

Publication Type: Conference contribution

Publication year: 2022

Journal

Lecture Notes in Computer Science Springer Verlag

Publisher: Springer Science and Business Media Deutschland GmbH

Book Volume: 13573 LNCS

Pages Range: 37-46

Conference Proceedings Title: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

Event location: Singapore, SGP

ISBN: 9783031185229

DOI: 10.1007/978-3-031-18523-6_4

Abstract

In federated learning for medical image analysis, the safety of the learning protocol is paramount. Such settings can often be compromised by adversaries that target either the private data used by the federation or the integrity of the model itself. This requires the medical imaging community to develop mechanisms to train collaborative models that are private and robust against adversarial data. In response to these challenges, we propose a practical open-source framework to study the effectiveness of combining differential privacy, model compression and adversarial training to improve the robustness of models against adversarial samples under train- and inference-time attacks. Using our framework, we achieve competitive model performance, a significant reduction in model’s size and an improved empirical adversarial robustness without a severe performance degradation, critical in medical image analysis.

Involved external institutions

Imperial College London / The Imperial College of Science, Technology and Medicine GB United Kingdom (GB) Technische Universität München (TUM) DE Germany (DE)

How to cite

APA:

Usynin, D., Klause, H., Paetzold, J.C., Rueckert, D., & Kaissis, G. (2022). Can Collaborative Learning Be Private, Robust and Scalable? In Shadi Albarqouni, Spyridon Bakas, Sophia Bano, M. Jorge Cardoso, Bishesh Khanal, Bennett Landman, Xiaoxiao Li, Chen Qin, Islem Rekik, Nicola Rieke, Holger Roth, Daguang Xu, Debdoot Sheet (Eds.), Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (pp. 37-46). Singapore, SGP: Springer Science and Business Media Deutschland GmbH.

MLA:

Usynin, Dmitrii, et al. "Can Collaborative Learning Be Private, Robust and Scalable?" Proceedings of the 3rd MICCAI Workshop on Distributed, Collaborative, and Federated Learning, DeCaF 2022, and the 2nd MICCAI Workshop on Affordable AI and Healthcare, FAIR 2022, held in conjunction with the 25th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2022, Singapore, SGP Ed. Shadi Albarqouni, Spyridon Bakas, Sophia Bano, M. Jorge Cardoso, Bishesh Khanal, Bennett Landman, Xiaoxiao Li, Chen Qin, Islem Rekik, Nicola Rieke, Holger Roth, Daguang Xu, Debdoot Sheet, Springer Science and Business Media Deutschland GmbH, 2022. 37-46.

BibTeX: Download